There has been little research conducted to determine if pure invasive ductal carcinoma (IDC) with no concomitant ductal carcinoma in situ (DCIS) differs from mixed IDC with DCIS in etiology, prognosis, or microenvironment, although there are some suggestions in the literature that pure IDC is likely to have a more aggressive course. The goal of this proposed research is to conduct a thorough examination of this phenomenon. We will examine risk factors for pure IDC vs mixed IDC with DCIS using data from a case- control study (n= 3715) designed to examine predictors of early/aggressive breast cancer in African-American and European-American women. We will first determine if pure IDC is more common in AA than in EA women, and characterize associated tumor characteristics for both phenotypes. Risk factors, particularly reproductive and hormonal factors, will be examined, with consideration of factors related to socioeconomic status and screening. Then, in a second cohort of 1588 women treated at Roswell Park Cancer Institute from 1995 to 2005 and followed up for treatment outcomes, we will examine associations between tumor and disease characteristics and pure IDC as well as mixed IDC with DCIS. Cox proportional hazard modeling will be used to determine disease and treatment outcomes among women in both of these groups, with consideration of other disease characteristics and treatments received. We hypothesize that women with pure IDC will be more likely to have aggressive tumor characteristics, and have more recurrences and shorter survival times than women with mixed IDC with DCIS. Finally, in an exploratory analysis using banked frozen tissue samples, we will microdissect stromal tissue from tumors from 15 women with pure IDC and 15 women with mixed IDC with DCIS and use real-time quantitative PCR to examine levels of expression of IL-6, TNF-, IFN-?, Cox-2, IGF-1, bFGF, TGF-1, fmsFLT1, MMP11 and TIMP-3, known for their role in tumor development, within each group. This proposed multidisciplinary research study, led by an epidemiologist and a breast biologist, will be the first to investigate the distribution f pure IDC and mixed IDC with DCIS by ancestry and to examine risk factors for each pathway, a previously unexplored area. In a large data set, we will evaluate the prognostic significance of these clinical subgroups, and in an exploratory pilot analysis, we will examine mechanisms by which specific pathways in the stromal microenvironment contribute to the pathobiological differences in IDC with and without concomitant DCIS.